A change to the privacy policy of X, formerly Twitter, reveals that posts made on the social network will be used to train the company’s “artificial intelligence models.” The social network does not have an AI of its own, but Elon Musk, owner of the platform, announced in July the creation of x.AI, his artificial intelligence company.
Soon, the training done with tweets/posts on the platform will be used to train the language model of the sister company. Despite the very similar name between the social network and the AI company, they are different legal entities. On the x.AI’s website, there is a text underscoring this separation, but assuring that it works “very closely” with Musk’s other companies to achieve its mission of “unlocking the secrets of the universe.”
The change in the privacy policy is in item 2.1 of X/Twitter, which deals with the use of “information to provide and operate products and services on the social network.”
By using X/Twitter’s post history, the x.AI is left with the knife and cheese in hand to acquire content for their training — even if there are tweets of questionable veracity. In theory, only Twitter has access to these posts. In reality, the platform continues with bots on the loose.
In April, Elon Musk, the owner of Xwitter and x.AI, threatened to sue Microsoft for training its artificial intelligence with tweets. The threat, which was yet another “attack” by the rival having removed X from its social media management service, shows that the problem (for all these companies) is when a competitor takes your data to get away with it.
Musk also follows a campaign to improve long-form publishing tools on the platform, aimed at bringing journalistic content to Xwitter — and using it to train the x.AI’s LLM. It’s the classic: if something is free, the product is you.
Companies are changing policies to “authorize” AI training
It’s not just the social network formerly known as Twitter that has changed its privacy policy to authorize AI training. Zoom, Brave (browser that “focuses on your privacy”), Microsoft and AWS have recently changed their policies to allow some level of data use by artificial intelligences.
